def try_merge(edition, ekey, thing):
thing_type = thing['type']['key']
if 'isbn_10' not in edition:
print(edition)
asin = edition.get('isbn_10', None) or edition['asin']
if 'authors' in edition:
authors = [i['name'] for i in edition['authors']]
else:
authors = []
a = amazon_merge.build_amazon(edition, authors)
assert isinstance(asin, six.string_types)
assert thing_type == '/type/edition'
#print edition['asin'], ekey
if 'source_records' in thing:
if 'amazon:' + asin in thing['source_records']:
return True
return source_records_match(a, thing)
#print 'no source records'
mc = get_mc(ekey)
#print 'mc:', mc
if mc == 'amazon:' + asin:
return True
if not mc:
return False
data = get_from_local(mc)
e1 = build_marc(fast_parse.read_edition(data))
return amazon_merge.attempt_merge(a, e1, threshold, debug=False)
def try_merge(edition, ekey, thing):
thing_type = thing['type']['key']
if 'isbn_10' not in edition:
print edition
asin = edition.get('isbn_10', None) or edition['asin']
if 'authors' in edition:
authors = [i['name'] for i in edition['authors']]
else:
authors = []
a = amazon_merge.build_amazon(edition, authors)
assert isinstance(asin, basestring)
assert thing_type == '/type/edition'
#print edition['asin'], ekey
if 'source_records' in thing:
if 'amazon:' + asin in thing['source_records']:
return True
return source_records_match(a, thing)
#print 'no source records'
mc = get_mc(ekey)
#print 'mc:', mc
if mc == 'amazon:' + asin:
return True
if not mc:
return False
data = get_from_local(mc)
e1 = build_marc(fast_parse.read_edition(data))
return amazon_merge.attempt_merge(a, e1, threshold, debug=False)
def load_part(archive_id, part, start_pos=0):
print 'load_part:', archive_id, part
global rec_no, t_prev, load_count
full_part = archive_id + "/" + part
f = open(rc['marc_path'] + "/" + full_part)
if start_pos:
f.seek(start_pos)
for pos, loc, data in read_marc_file(full_part, f, pos=start_pos):
rec_no += 1
if rec_no % chunk == 0:
progress(archive_id, rec_no, start_pos, pos)
if is_loaded(loc):
continue
want = ['001', '003', '010', '020', '035', '245']
try:
index_fields = fast_parse.index_fields(data, want)
except KeyError:
print loc
print fast_parse.get_tag_lines(data, ['245'])
raise
except AssertionError:
print loc
raise
if not index_fields or 'title' not in index_fields:
continue
edition_pool = pool.build(index_fields)
if not edition_pool:
yield loc, data
continue
rec = fast_parse.read_edition(data)
e1 = build_marc(rec)
match = False
seen = set()
for k, v in edition_pool.iteritems():
for edition_key in v:
if edition_key in seen:
continue
thing = None
while not thing or thing['type']['key'] == '/type/redirect':
seen.add(edition_key)
thing = withKey(edition_key)
assert thing
if thing['type']['key'] == '/type/redirect':
print 'following redirect %s => %s' % (edition_key, thing['location'])
edition_key = thing['location']
if try_merge(e1, edition_key, thing):
add_source_records(edition_key, loc, thing, data)
match = True
break
if match:
break
if not match:
yield loc, data
def marc_match(e1, loc):
rec = fast_parse.read_edition(get_from_local(loc))
try:
e2 = build_marc(rec)
except TypeError:
print rec
raise
return attempt_merge(e1, e2, threshold, debug=False)
def get_record(key, mc):
data = get_from_archive(mc)
try:
rec = fast_parse.read_edition(data)
except (fast_parse.SoundRecording, IndexError, AssertionError):
print mc
print edition_key
return False
try:
return marc.build_marc(rec)
except TypeError:
print rec
raise
def get_record(key, mc):
data = get_from_archive(mc)
try:
rec = fast_parse.read_edition(data)
except (fast_parse.SoundRecording, IndexError, AssertionError):
print(mc)
print(key)
return False
try:
return marc.build_marc(rec)
except TypeError:
print(rec)
raise
def get_marc(loc):
try:
filename, p, l = loc.split(':')
except ValueError:
return None
if not os.path.exists(marc_path + filename):
return None
f = open(marc_path + filename)
f.seek(int(p))
buf = f.read(int(l))
f.close()
rec = fast_parse.read_edition(buf)
if rec:
return build_marc(rec)
def get_marc(loc):
try:
filename, p, l = loc.split(':')
except ValueError:
return None
if not os.path.exists(marc_path + filename):
return None
f = open(marc_path + filename)
f.seek(int(p))
buf = f.read(int(l))
f.close()
rec = fast_parse.read_edition(buf)
if rec:
return build_marc(rec)
def load_part(archive_id, part, start_pos=0):
print('load_part:', archive_id, part)
global rec_no, t_prev, load_count
full_part = archive_id + "/" + part
f = open(rc['marc_path'] + "/" + full_part)
if start_pos:
f.seek(start_pos)
for pos, loc, data in read_marc_file(full_part, f, pos=start_pos):
rec_no += 1
if rec_no % chunk == 0:
progress(archive_id, rec_no, start_pos, pos)
if is_loaded(loc):
continue
want = ['001', '003', '010', '020', '035', '245']
try:
index_fields = fast_parse.index_fields(data, want)
except KeyError:
print(loc)
print(fast_parse.get_tag_lines(data, ['245']))
raise
except AssertionError:
print(loc)
raise
if not index_fields or 'title' not in index_fields:
continue
edition_pool = pool.build(index_fields)
if not edition_pool:
continue
rec = fast_parse.read_edition(data)
e1 = build_marc(rec)
match = False
seen = set()
for k, v in edition_pool.iteritems():
for edition_key in v:
if edition_key in seen:
continue
seen.add(edition_key)
thing = withKey(edition_key)
assert thing
if try_merge(e1, edition_key, thing):
add_source_records(edition_key, loc, thing)
match = True
if not match:
yield loc, data
def get_ia(ia):
# read MARC record of scanned book from archive.org
# try the XML first because it has better character encoding
# if there is a problem with the XML switch to the binary MARC
xml_file = ia + "_marc.xml"
loc = ia + "/" + xml_file
for attempt in range(3):
if os.path.exists(xml_path + xml_file):
f = open(xml_path + xml_file)
else:
f = urlopen_keep_trying(base + loc)
if f:
try:
return loc, read_xml.read_edition(f)
except read_xml.BadXML:
pass
except xml.parsers.expat.ExpatError:
print 'XML parse error:', base + loc
pass
sleep(2)
url = base + ia + "/" + ia + "_meta.mrc"
try:
f = urlopen_keep_trying(url)
except urllib2.URLError:
pass
if not f:
return None, None
data = f.read()
length = data[0:5]
loc = ia + "/" + ia + "_meta.mrc:0:" + length
if len(data) == 0:
print 'zero length MARC for', url
return None, None
if 'Internet Archive: Error' in data:
print 'internet archive error for', url
return None, None
try:
return ia, fast_parse.read_edition(data, accept_electronic = True)
except (ValueError, AssertionError):
print `data`
raise
def marc_match(a, loc):
assert loc
rec = fast_parse.read_edition(get_from_local(loc))
e1 = build_marc(rec)
#print 'amazon:', a
return amazon_merge.attempt_merge(a, e1, threshold, debug=False)
def marc_match(a, loc):
assert loc
rec = fast_parse.read_edition(get_from_local(loc))
e1 = build_marc(rec)
#print 'amazon:', a
return amazon_merge.attempt_merge(a, e1, threshold, debug=False)
if not a:
return False
try:
return amazon.attempt_merge(a, e1, threshold, debug=False)
except:
print a
print e1
print thing['key']
raise
print 'mc:', mc
try:
assert not mc.startswith('ia:')
data = get_from_local(mc)
if not data:
return True
rec2 = fast_parse.read_edition(data)
except (fast_parse.SoundRecording, IndexError, AssertionError):
print mc
print edition_key
return False
except:
print mc
print edition_key
raise
if not rec2:
return False
try:
e2 = build_marc(rec2)
except TypeError:
print rec2
raise
